To determine whether the Oracle GoldenGate process is operating correctly, the fault monitor for the Oracle GoldenGate resource type probes these resources periodically.
The probe checks if the manager process is in the process table. If this is true, then the probe opens a socket to the manager port. If not, then the probe declares the resource as faulted.
If the manager is OK, the probe checks if there are extract or replicat processes running. If not, it checks the status of these processes with the following command:
# su - ggowner -c "echo 'info all' | base directory/ggsci"
If the status of every configured extract or replicat process is abended, the probe returns with the value of the property DEGRADED_CODE.
The number of tolerated consecutive DEGRADED_CODE probe return within the retry_interval seconds is calculated by the number 100 divided by DEGRADED_CODE. If this number is greater than the retry_interval divided by thorough_probe_interval, then the probe return values of DEGRADED_CODE will be tolerated forever.